Hb A₂ Episkopi – a novel δ-globin chain variant [HBD:c.428C>T] in a family of mixed Cypriot–Lebanese descent

ABSTRACT

Objectives: Thalassaemia is a potentially lethal inherited anaemia, caused by reduced or absent synthesis of globin chains. Measurement of the minor adult haemoglobin Hb A₂, combining α- with δ-globin, is critical for the routine diagnosis of carrier status for α- or β-thalassaemia. Here, we aim to characterize a novel δ-globin variant, Hb A₂ Episkopi, in a single family of mixed Lebanese and Cypriot ancestry with mild hypochromic anaemia and otherwise normal globin genotype, which also presents with a coincidental 0.78-Mb sequence duplication on chromosome 1 (1q44) and developmental abnormalities.

Methods: Analyses included comprehensive haematological analyses, cation-exchange high-performance liquid chromatography (CE-HPLC), cellulose acetate electrophoresis (CAE), Sanger sequencing and structure-based stability predictions for Hb A₂ Episkopi.

Results: The GCT > GTT missense mutation, underlying Hb A₂ Episkopi, HBD:c.428C > T, introduces a cd142 codon change in the mature protein, resulting in reduced normal Hb A₂ amounts and a novel, less abundant Hb A₂ variant (HGVS: HBD:p.A143V), detectable as a delayed peak by CE-HPLC. The latter was in line with structure-based stability predictions, which indicated that the substitution of a marginal, non-helical and non-interface residue, five amino acids from the δ-globin chain carboxy-terminus, was moderately destabilizing.

Discussion: Detection of the new variant depends on the diagnostic set-up and had failed by CAE and on an independent CE-HPLC system, which, in unfavourable circumstances, may lead to misdiagnoses of β-thalassaemia as α-thalassaemia. Given the mixed background of the affected family, the ethnic origin of the mutation is unclear, and this study thus suggests awareness for possible detection of Hb A₂ Episkopi in both the Cypriot and the Lebanese populations.

KEYWORDS:

• Haemoglobin A₂
• globin variant
• thalassaemia
• cation-exchange HPLC
• cellulose acetate electrophoresis

Introduction

The potentially lethal monogenic anaemia β-thalassaemia requires lifelong and costly disease management [1]; therefore, national screening and prevention programmes are crucial to countries with a high proportion of disease carriers. The latter present haematologically with characteristically low mean cell volume and, through depletion of β-globin and thus the main adult haemoglobin Hb A (α₂β₂), show an increase in the minor adult haemoglobin Hb A₂ (α₂δ₂) from 2.5–3.0% to above 3.5% of total haemoglobin [2]. In consequence, δ-globin expression is central to the haematological diagnosis of thalassaemias. This is exemplified by the situation in which double heterozygotes for δ- and β-thalassaemia present with an atypical phenotype characterized by microcytosis and hypochromia, but with normal Hb A₂, similar to certain α-thalassaemia heterozygotes. In countries such as Cyprus, where the occurrence of α-, β- and δ-thalassaemia mutations is common [3], the presence of δ-thalassaemia masks the typical β-thalassaemia phenotype and thus interferes with population screening for β-thalassaemia carriers [4]. Cases with borderline or abnormal haematological findings are, therefore, routinely undergoing molecular analysis of the δ-globin gene as well. This allows a definitive diagnosis and safeguards disease prevention and national disease control for thalassaemia on the one hand, while on the other hand continuously accumulating and updating a comprehensive catalogue of mutations in the respective population [5]. Here, we report a novel δ-globin gene variant, Hb A₂[δ^A142V] or Hb A₂ Episkopi, unresolved by cellulose acetate electrophoresis (CAE) but detectable by cation-exchange high-performance liquid chromatography (CE-HPLC), presenting in a normal genetic background of a family with mixed Lebanese–Cypriot descent.

Materials and methods

Subjects

Study subjects were analyzed between 2011 and 2013 as part of the national thalassaemia prevention programme of the Republic of Cyprus and have given full informed consent for their inclusion in this study.

Haematological analyses

Haematological studies were performed using routine methods. Haemoglobin analysis and detection of Hb A₂ Episkopi were achieved by CE-HPLC on the β-thalassaemia VARIANT™ program (Bio-Rad Life Sciences), while employment of an alternative HPLC system (Shimadzu) (see online supplementary material Fig. S1) and of standard CAE procedures and equipment (Helena Laboratories) had failed to do so. Additional post-analysis HPLC peak quantification was performed using FiJi ImageJ 1.49 [6].

Molecular analyses

Nucleotide and amino-acid numbers are given with RefSeq NG_000007 as reference for genomic DNA positions and NM_000519.3 as reference for human HBD cDNA and predicted amino-acid sequences. By convention in the haemoglobin field and contrary to guidelines of the Human Genome Variation Society (HGVS), amino-acid references do not count the initiating methionine, absent in the mature protein, and therefore differ by −1 from corresponding HGVS position references.

In addition to the analyses in the δ-globin gene HBD, the specific study subjects were tested for –α^3.7, ααα, α^–5ntα, αAgrα, α^PolyA1α and α^PolyA2α mutations of the α-globin locus [7] and for any mutations in the β-globin locus between 288 nt upstream of the HBB start codon and 240 nt downstream of the HBB stop codon [3]. Mutations in HBD were detected by sequencing of two overlapping PCR fragments. The first fragment of 1161 bp was amplified with primers δF1: 5′-TACATTCCACTATATTAGCC-3′ and δ6R: 5′ CAGTATTCTATGCCTCTCAT-3′ and spanned the 5′ UTR, exon 1 and exon 2, extending from −328 to +814 relative to the HBD start codon. The second fragment of 947 bp, amplified with primers δ5F: 5′-TGCATACCAGCTCTCACCTG-3′ and δ8R: 5′-CAGGAACCTTCTTACACACC-3′, extended from +707 to +1653 relative to the HBD start codon. Sequencing reactions (Invitrogen, BigDye Terminator v. 1.1 cycle sequencing kit) used primers δ1F, δ15S (5′-AACCAACCTGCTCACTGGAG-3′) and δ17S (5′-ATTTATGCTGATGGGAATAAC-3′) and were analyzed on a 3130xl Genetic Analyzer (Applied Biosystems Hitachi). Hb Episkopi has been deposited in the IthaGenes database with IthaID 2560 (www.ithanet.eu) [8].

Bioinformatics and statistical analyses

In silico protein-structure and stability analyses were based on RCSB PDB entry 1si4 of the relaxed (R2) state of normal Hb A₂ [9]. Modelling of normal and variant Hb A₂ was performed by FoldX3.0 in command-line mode [10] for the creation and energy-optimization of variants, and using YASARA [11] and the POV-Ray rendering tool (www.povray.org) for rendering of images. Structure-dependent comparative energy calculations for tetramers by FoldX assumed default conditions, including a temperature of 37°C and a pH of 7, and used five iterations to calculate averages and standard deviations of energy differences (kcal mol^-1) between variant and normal Hb A₂. The significance of changes in Hb A₂ was assessed for each test case against a reference population of individuals matched for gender and α- and β-globin genotype, using the Microsoft Excel NORM.DIST function. The isoelectric point (pI) for all δ-globin variants and pI differences for variants compared to normal δ-globin are based on the protein pI estimate of the Isoelectric Point Calculator (IPC) [12].

Results

An index case presenting with anatomical and haematological abnormalities

The index case (M1) from Episkopi of the Limassol district of Cyprus was first seen at the age of 1 month for evaluation regarding cryptorchidism and a U-shaped cleft palate. The proband had, in addition, deep-set, hypoplastic toe-nails, but otherwise his evaluation was unremarkable. Review of the family history revealed that the child was born to healthy, non-consanguineous parents of mixed Cypriot–Lebanese descent, with a family history non-contributory for clefting but contributory on the maternal side for cryptorchidism.

These findings indicated more comprehensive analyses, which revealed an apparently normal male karyotype, normal biochemistry, ophthalmology, cardiology and ultrasound scan of the abdomen, while array-CGH revealed a maternally inherited duplication of 0.78 Mb, on chromosome 1 on chromosomal band 1q44, location: 244,935,100-245,724,786 [build GRCh37 (hg19)] [arr 1q44(244,935,100-245,724,786)x3 mat], the clinical significance of which remains unclear. The father was unavailable for testing.

Subsequent analyses of blood smears and a duplicate complete blood count (CBC) for the index case at 10 months of age and for his mother (F1), aged 33, in both cases revealed mild hypochromic, microcytic anaemia. Mean cell volume (73.3 fl and 78.5 fl) and mean cell haemoglobin (24.6 pg and 26.7 pg) for M1 and F1, respectively, were slightly below the reference range, while other CBC parameters and indices for iron deficiency (zinc protoporphyrin and RDW-CV) were in the normal range (Table 1).

Table 1. Study subjects.

Subject	Place of residence	Zpp (moles Zp/mole haeme)	RBC (10^6/μl)	Hb (g/dl)	MCV (fl)	MCH (pg)	Hb A2 (CAE) (%)	Genotype
								α-globin gene	β-globin gene
M1	Episkopi	74	4.4	11.1	72.7	25.0	2.6	αα/αα	Ν/N
Reference		<90	3.7–5.3	10.5–13.5	70–86	23–31	1.6–3.3
F1	Episkopi	102	4.84	12.9	78.5	26.7	1.9*	αα/αα	N/N
Reference		<105	4.0–5.2	12.0–16.0	80–100	26–34	2.2–3.5

Two individuals from the Cypriot family of Cypriot–Lebanese ancestry are heterozygous for Hb A₂ Episkopi.

Reference ranges are normal ranges for children of 0.5–2 years and for adult women, respectively [20].

αα/αα, N/N – diagnosed homozygous normal according to haematological sample parameters, respectively, for the α-globin locus and for the β-globin locus after the detection assays specified in the section ‘Materials and methods’.

*In agreement with literature references [20], Hb A₂ levels for F1 are slightly but significantly below gender-matched control levels of the respective reference population in Cyprus (p = 0.033), in contrast to those for M1 (p = 0.495). The underlying cause is unknown.

CAE analysis showed reduced Hb A₂ for F1 but normal levels for M1 (Table 1), each compared to the age-matched (and for the mother gender-matched) reference range. However, CE-HPLC analysis revealed a split peak for Hb A₂, indicating a novel stable Hb A₂ variant (with 87% the peak area of normal Hb A₂) and prompting molecular analyses.

Confirmation of the novel δ-globin variant by sequence analysis

Comprehensive globin sequence analysis and gap-PCR [13] of the index case M1 revealed heterozygosity for a hitherto unreported δ-globin mutation (Figure 1(a)), in the absence of any of the other α- or β-globin mutations tested, with identical findings for F1. Mutations in the globin loci are, therefore, not responsible for the slightly abnormal or borderline haematology of M1 and F1. The novel δ-globin missense mutation introduces a codon change GCT > GTT (alanine to valine) at codon 143 of the HBD gene (and thus at amino acid 142 of the mature δ-globin protein). For the same nucleotide position, an alternative GCT > GAT mutation has been characterized before and brings about an amino acid change of alanine to aspartate, giving rise to HbA₂ Fitzroy instead [14].

Figure 1. Characterization of the Hb A₂ Episkopi mutation (a) Sequencing revealed a GCT > GTT codon change (A143V, indicated by an arrow) for δ-globin underlying the observation of a novel variant. (b) CE-HPLC analysis reveals a split peak at the Hb A₂ position (expected elution time 2.9 minutes), with a trailing peak (elution time 3.07 minutes) of comparable height to the normal Hb A₂, indicating a variant of similar stability. The inlayed cake diagram shows the relative abundance of each Hb A₂ species. (c) The known Hb A₂ tetramer structure (PDB 1si4 [9], left panel) has been energy-optimized using FoldX to accommodate the A143V amino-acid change (right panel), which is located five amino acids from the carboxy-terminus of δ-globin. For a comparison of the most common Cypriot variants, see online supplementary material Fig. S2. The α- and δ-globin chains (α1, α2 and δ1, δ2) are displayed as flat ribbon diagrams in grey turquoise and yellow, respectively, with residue V142 in each mature δ-globin chain highlighted in mauve and displayed as a ball-and-stick diagram. Numbers indicate the δ-globin C and N termini and the variant amino-acid. Haeme groups are shown as stick diagrams in red.

Predicted and observed consequences of the amino-acid substitution

HBD:c.428C > T is the mutation closest to the 3′ end of the δ-globin gene yet reported in Cyprus [15], positioning the affected amino-acid at the transition of the final α-helix of δ-globin towards the flexible carboxy-terminus. The predicted A > V amino-acid change from one non-polar, uncharged aliphatic side chain to another maintains the predicted isoelectric point (pI) of 7.42 for the resultant variant δ-globin (see Table 2) [12], in contrast to the HBD:c.428C > A mutation underlying HbA₂ Fitzroy [14], which results in an A > D amino-acid change and a predicted pI of 6.87. This is in line with an inability to detect the new Hb A₂[δA142V] variant by the charge-dependent method of CAE in the basic heterozygote state, necessitating detection of a split peak by CE-HPLC (Figure 1(b)) to avoid possible misdiagnosis, while the more acidic HbA₂ Fitzroy is readily detectable by CAE [14]. The properties of both alternative same-residue variants are compared in Table 3. Of note, elevated oxygen affinity was inferred for Hb Fitzroy from the equivalent Hb A variant Hb Ohio (Hb A[βA142D]) [16], and we likewise infer for Hb Episkopi that it will have elevated oxygen affinity, as was shown for its paralogous β-globin variant Hb Waterland (Hb A[βA142V]) [17]. For the Hb A₂ Episkopi heterotetramer, structure-based stability prediction gives only subtle changes in the predicted protein backbone of the heterotetramer but indicates a moderately destabilizing effect of the mutation (see Figure 1(c) and Table 2), which agrees with the observed steady-state abundance of the new variant being slightly below that of normal Hb A₂ in the heterozygote state_.

Table 2. Stability of common Hb A₂ variants detected in Cyprus.

δ-globin variant	Codon change	Common name	HGVS protein name	ΔΔG	ΔpI	dbSNP reference	IthaGenes reference	Variant origin and reference
Normal	n/a	n/a	HBD	0	0	n/a	GeneID 9	n/a
T4I	ACT > ATT	CD4 (ACT > ATT)	HBD:p.T5I	1.13 ± 0.53	0	rs35406175	1331	Greek, Cypriot [4,22]
N19K	AAC > AAA	Hb A2 Famagusta	HBD:p.N20K	−1.53 ± 0.07	+0.41	n/a	2292	Cypriot [15]
G25D	GGT > GAT	Hb A2 Yokoshima	HBD:p.G26D	8.13 ± 1.19	−0.553	rs34389944	1360	Japanese, Cypriot [22]
A27S	GCC > TCC	Hb A2 Yialousa	HBD:p.A28S	1.40 ± 0.01	0	rs35152987	1332	Greek, Cypriot, Sardinian [4,22]
V98M	GTG > ATG	Hb A2 Wrens	HBD:p.V99M	−0.94 ± 0.07	0	rs28933076	1339	Black [22,23]
R116C	CGC > TGC	Hb A2 Troodos	HBD:p.R117C	3.80 ± 0.03	−0.635	rs33971270	1384	Cypriot, Greek [4,22]
L141P	CTG > CCG	Hb A2 Pelendri	HBD:p.L142P	17.33 ± 0.42	0	rs33956485	1392	Cypriot [4,22,24]
A142V	GCT > GTT	Hb A2 Episkopi	HBD:p.A143V	3.50 ± 0.38	0	n/a	2560	Cypriot–Lebanese [this report]

Predicted free energy changes (ΔΔG) for variant vs. normal Hb A₂ tetramers were calculated using FoldX 3.0 and five modelling iterations per variant, with negative scores indicating a stabilizing mutation [21]. The change in isoelectric point (ΔpI) is based on predictions using the Isoelectric Point Calculator (IPC) [12].

Table 3. Comparison of properties for Hb A₂ Episkopi and Hb A₂ Fitzroy.

	Hb A2 Episkopi (HBD:p.A143V)	Hb A2 Fitzroy (HBD:p.A143D)
Ratio Hb A2/Hb A2’ in heterozygotes	0.87 (CE-HPLC)	1 (CAE)
Paralogous Hb A (β-globin) variant	Hb Waterland	Hb Ohio
Inferred change in oxygen affinity	elevated	elevated
Codon change in variant δ-globin	GCT > GTT	GCT > GAT
pI of variant δ-globin ^[IPC]	7.422	6.869^IPC
pI difference compared to normal δ-globin^[IPC]	0	−0.553
Charge of variant δ-globin at pH 7.4 (cytosol) ^[IPC]	0	−1
Molecular mass [Da] of variant δ-globin ^[IPC]	16083.53	16099.49
Mass difference [Da] compared to normal δ-globin^[IPC]	+28.05	+44.01
dbSNP reference	n/a	rs35848600
IthaGenes reference	2560	1393
Variant origin	Cypriot–Lebanese	Greek

Known and predicted properties of the same-residue δ-globin variants Hb A₂ Episkopi (HBD:p.A143V) and Hb A₂ Fitzroy (HBD:p.A143D) are summarized as given in this study for the former and in the original publication for the latter [14], and as calculated by the Isoelectric Point Calculator (^IPC) [12]. No experimental data exist for the oxygen affinity of either variant.

Discussion

Here, we have described and sequenced a novel δ-globin variant in a family of mixed Cypriot–Lebanese descent and named after the family’s place of residence, Hb A₂ Episkopi. Experimental CE-HPLC data and stability predictions suggest that the variant is slightly destabilized compared to normal Hb A₂ and is readily detectable as an additional peak (albeit without baseline separation) on suitable HPLC systems.

This description of Hb A₂ Episkopi increases the heterogeneity of the δ-globin gene in Cyprus from 15 to 16 alleles [3], including 11 variants. Of routine laboratory methods, CE-HPLC is able to discern Hb A₂[δA142V] from normal Hb A₂ where a suitable CE-HPLC system is available, owing to the variant’s stability and its overall charge, while CAE fails to detect the variant. This is in contrast to HbA₂ Fitzroy (Hb A₂[δA142D]), which results from an alternative missense mutation of the same codon and is readily detectable by CAE [14]. Our observations once more emphasize the need to employ multiple analysis methods for the routine detection of haemoglobin variants, which also applies to environments without access to HPLC equipment, where complementary electrophoresis methods might be used instead. For instance, the combination of methods detecting changes in pI or in charge at a given pH, such as isoelectric focussing (IEF) or CAE [18,19], could be combined with methods detecting changes in molecular weight, such as acid-urea-Triton X-100 polyacrylamide gel electrophoresis (AUT-PAGE) [19], to minimize false-negative variant detection and thus the risk of misdiagnosis of haemoglobinpathies.

With this paper, we have established the detection and identity of the novel δ-globin variant Hb A₂ Episkopi, which may lead to the targeted discovery of the same sequence variant in other, and in particular Lebanese, populations.

Acknowledgements

The authors thank all study subjects for their participation, the Population Thalassaemia Screening Laboratory of the Thalassaemia Centre of Cyprus for its services and Eleni Karitzi for technical assistance.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Dr Carsten W. Lederer is a group leader in the Department of Molecular Genetics Thalassaemia at the Cyprus Institute of Neurology and Genetics and a Lecturer at the Cyprus School of Molecular Medicine, Nicosia, Cyprus. He is an expert in the field of haemoglobinopathies and in the gene therapy of the haematopoietic system.

Ms Eleni Pavlou is a laboratory scientific officer in the Department of Molecular Genetics Thalassaemia at the Cyprus Institute of Neurology and Genetics. She is an expert in the molecular diagnosis of thalassaemias.

Dr George A. Tanteles MD DM is Head of the Clinical Genetics Clinic and a Consultant Clinical Geneticist at the Cyprus Institute of Neurology and Genetics.

Dr Paola Evangelidou is a scientist in the Department of Cytogenetics and Genomics at the Cyprus Institute of Neurology and Genetics. She is an expert in the field of cytogenetic diagnosis.

Dr Carolina Sismani is Head of the Department of Cytogenetics and Genomics at the Cyprus Institute of Neurology and Genetics and Assistant Professor at the Cyprus School of Molecular Medicine, Nicosia, Cyprus. She is an expert in the field of cytogenetic diagnosis.

Dr Annita Kolnagou is a Consultant Physician, Head of the Thalassaemia Clinic Paphos, Cyprus.

Dr Maria Sitarou MD, MSc in Haemoglobinopathy, is Head of the Thalassaemia Clinic General Hospital Larnaca, Cyprus.

Dr Soteroulla Christou is Head of the Thalassaemia Clinic Nicosia, Cyprus.

Dr Michael Hadjigavriel is Head of the Thalassaemia Clinic Limassol, Cyprus. He is a specialist in internal medicine, and an associate of the course MSc in Haemoglobinopathy.

Dr Marina Kleanthous is Head of the Department of Molecular Genetics Thalassaemia at the Cyprus Institute of Neurology and Genetics and an Associate Professor at the Cyprus School of Molecular Medicine, Nicosia, Cyprus. She is an expert in the field of haemoglobinopathies and of thalassaemia in particular.

ORCID

Carsten W. Lederer http://orcid.org/0000-0003-3920-9584

George A. Tanteles http://orcid.org/0000-0002-8117-9852

Additional information

Funding

This work was, in part, funded by the European Union Seventh Framework Programme THALAMOSS project [Project ID: 306201].